A Brain Computer Interface (BCI) is the generic term used to describe any kind of system that serves as a communication bridge between the brain (human or not) and an artificial module. It’s a field of research in which wide investment has been made since the 1970’s, especially in the clinical fields and ergonomics. Generally speaking, any kind of brain activity that can be recorded can be used as a means of communicating with another system. Through the use of statistical classification techniques it’s possible to associate certain states or characteristics of the recorded signal – which the experiment subject learns to control - to any procedure, usually mediated by a computer.
Many techniques have been developed to help us look and better understand the way the brain works. They range from imaging techniques (like MRI, fMRI, fNIRS or PET), to electrophysiological ones (like EEG, EcG or MEG). While the first category is usually used to obtain high resolution images of brain structures and the second one to register and analyze the electrical activity produced by the brain, with a high temporal resolution – which is why they are the ones mainly used in the field of BCI’s. In pair with such methods, although a different area in itself, includes brain implants capable of communicating directly with the neuronal tissue - neuroprosthetics.
Of all the different means avaliable, the registering of the electroencephalographic (EEG) activity is the most developed and extensively researched of this fields. It allows us, in a non-invasive way, to peak the brain functioning with a high temporal resolution – furthermore, it is now well established that different brain states produce distinct observable activity. With the help of electrodes placed on the scalp, it is possible to feed this activity and their respective variations and patterns to any system capable of classifying and detecting them in real time and act accordingly (making this a field highly interconnected to that of machine learning)....